THE OPTIMAL INSULATION DETAIL OF THE THERMAL BRIDGE ADJACENT TO HOT WATER PIPES IN APARTMENT BUILDING SLABS. Seung Yeong Song & Kwang Woo Kim

Transcription

1 THE OPTIMAL INSULATION DETAIL OF THE THERMAL BRIDGE ADJACENT TO HOT WATER PIPES IN APARTMENT BUILDING SLABS Seung Yeong Song & Kwang Woo Kim Department of Architecture, Seoul National University, San 56-1, Shinlim-Dong, Kwanak-Gu, Seoul, , Korea ABSTRACT The methods for evaluating the thermal performance of each insulation detail alternative with the multidimensional heat transfer simulation are presented to determine the optimal insulation details of the thermal bridges adjacent to hot water pipes in apartment building slabs. The optimal insulation detail of the side wall-slab joint is presented based on the evaluation of inside surface condensation and life cycle costs. details of the joints where hot water pipes for heating are and where multi-dimensional heat flow path occurs. Thus, this study aims to present the methods for evaluating the thermal performance of the insulation details of these joints, quantitatively, with the multidimensional heat transfer simulation. The optimal insulation detail of the side wall-slab joint is presented based on the evaluation of inside surface condensation and life cycle costs. INTRODUCTION There are many cases in which the layer of insulation is disconnected by structural components at the joints in apartment building envelope. These joints become thermal bridges at which the risk of inside surface condensation and heat loss increase. Especially at the joints adjacent to hot water pipes for heating (see Fig. 1), the heat loss is so great that the outside surface s are higher by 24& in the winter than those of surrounding regions. To improve the thermal performance of these joints, the optimal insulation details should be applied after the evaluation of each insulation detail alternative. If the optimal insulation details are applied to each joint, there is a good chance of energy conservation since apartment buildings are envelope load dominated buildings in which components with similar structures are repeated. In the existing insulation details of these joints, subsidiary insulation is installed to decrease heat loss. (see Fig. 1) However, there are some cases in which heat loss increases due to misplacement and wrong dimensions of subsidiary insulation. These are attributable to the lack of quantitative thermal performance evaluation methods for the insulation Fig. 1. Side wall-slab joint THERMAL PERFORMANCE EVALUATION METHODS As the thermal resistance of the joints increases, the risk of inside surface condensation in the winter and heating cost decrease. Additional benefits may be increased occupant comfort and reduced heating 1 1 There is no pre-installed central cooling system in Korean apartment buildings. 1

2 system capacity. Since these benefits are difficult to quantify and are likely to be negligible [1], the suggested evaluation methods focus on inside surface condensation and life cycle costs. INSIDE SURFACE CONDENSATION Boundary conditions Inside surface condensation occurs when inside surface is lower than inside air dew point. Since inside air dew point is determined by inside air and relative humidity, inside surface condensation may or may not occur at the same surface. The inside air and relative humidity are varied dramatically according to the location, heating method, extent of insulation, and way of life. Thus, it is necessary to set the standard inside air and relative humidity conditions for the evaluation. As the outside air drops, the inside surface will drop accordingly. Concerning outside air, the safest evaluation results can be obtained if the worst condition is applied. In this study, the standard inside air and relative humidity, and the lowest outside air of Seoul are determined on the basis of the conditions in KSF2295, Korean Standard on the Method of Testing for Condensation at Windows and Doors. (see Table 1) Table 1. The standard condition for the evaluation of inside surface condensation (KSF2295) Inside air Inside air relative humidity Inside air dew point Outside air (Seoul) 20& 50% 9.3& -15& On the inside surface of the joints, heat transfer coefficient decreases due to the drop in air speed. If the heat transfer coefficient on the inside surface decreases when inside surface is lower than inside air, the inside surface becomes lower. Thus, the risk of inside surface condensation increases. In this study, a heat transfer coefficient of 3.5 W/m 2 & is used on the inside surface. [2] Evaluation methods It can be determined whether the inside surface condensation may or may not occur by comparing the lowest inside surface (T l ) to the inside air dew point (9.3&) of the standard condition. The lowest inside surface (T l ) for the standard condition can be found by the steady state heat transfer analysis. Apartment buildings are residential buildings with continuous night time heating. Thus, the steady state heat transfer analysis is a safe analysis method to get the lowest inside surface. The lowest inside surface (T l ) can be represented as the dimensionless lowest inside surface (T dl ) as in Eq. (1). [3] If the lowest inside surface (T l ) is substituted with the inside air dew point (T c ) after the rearrangement of Eq. (1) to Eq. (2), the outside air at which the inside surface condensation will occur (T oc ) can be found. It is then possible to evaluate the inside surface condensation for any combination of the inside air (T i ), relative humidity, and outside air (T o ). T dl Tl To = Ti To (1) Outside air causing inside surface condensation, T oc ( o C) % Inside air : 20 o C 70% 80% 60% 50% 40% 30% 20% Inside air relative humidity 90% T o Tdl Ti Tl = «Tl => Tc (2) Tdl 1 The outside air at which the inside surface condensation will occur (T oc ) to the dimensionless lowest inside surface (T dl ) is presented in Fig. 2 when the inside air condition is 20& and 1090% (R.H.) Dimensionless lowest inside surface, T dl Fig. 2. Outside air causing inside surface condensation LIFE CYCLE COST Economic factors related with the thermal performance of the insulation detail are material, labor, cooling, heating, and maintenance costs. Life 2

3 cycle costs are the sum of these costs during the life time including construction period. If the life cycle costs trend of each insulation detail alternative can be found, it is possible to determine the economically optimal insulation detail of the joints. To calculate the life cycle costs of each insulation detail alternative, it is necessary to quantify the costs of all economic factors. However, there is no central cooling system in Korean apartment buildings. Occupants use electric fans or room air conditioners for cooling depending on individual choice. Thus, it is difficult to quantify the cooling cost due to the heat gain through the joints. According to the survey of management expenses of 1,896 apartment buildings in Korea, annual cooling cost in 1996 is only 4.7% of total expenses related to energy use. [4] So, it is expected that the cooling cost due to the heat gain through the joints is so small and negligible. Considering the occupants' convenience, insulation detail alternative that needs repair or replacement Fig. 3. Heat flow at the joints Table 2. Hot water supply schedule for heating 2 Daily lowest outside air (&) Heating time Hot water (&) below -15 4~7, 11~13, 17~ ~-10 3~6, 11~13, 17~ ~-5 4~6, 11~13, 17~ ~15 4~6, 17~ In Korean apartment buildings, intermittent heating is generally applied. during the life time cannot be taken into account. So, it is possible not to consider the maintenance cost. In this study, cooling and maintenance costs are not considered in calculating the life cycle costs of each insulation detail alternative. Each insulation detail alternative can be assumed to have the same labor cost, when there is no difference in the amount of work and the process of construction. If the same amount of cost is added to or subtracted from the life cycle costs of each insulation detail alternative, life cycle costs trend is not changed. In this study, labor cost is not considered in determining the economically optimal insulation detail when each alternative can be assumed to have the same cost. Heating cost Heating cost during the life time can be calculated by annual heating cost multiplied by present value coefficient. To calculate annual heating cost, annual heat loss through the joints should be known. To calculate the annual heat loss through the joints, multi-dimensional unsteady state heat transfer simulation is performed. To evaluate each insulation detail alternative correctly, annual heat loss through the joints should be calculated while applying actual hot water supply schedule for heating. Heat flow at the joints adjacent to hot water pipes in heating is presented in Fig. 3. Q1 and Q5 are heat gain and loss on the outside surface, respectively. Q2 is heat loss from the stored heat in envelope. Q3 and Q4 are heat gain and loss on the inside surface, respectively. Q6 and Q7 are heat gain and loss from hot water pipes, respectively. In this study, annual sum of Q4 and Q7 are defined as annual heat loss through the joints. Boundary for calculating annual heat loss is represented by the thick solid and dotted lines in Fig. 3. The generalized hot water supply schedule for heating in Korean apartment buildings is presented in Table 2. In calculating annual heat loss with the multidimensional unsteady state heat transfer simulation, it is inefficient to model the whole building envelope. To evaluate each insulation detail alternative, the region affected by the thermal bridge can be isolated from the remainder of the building envelope by identifying a set of adiabatic planes. [5] In this study, the effective calculation domain, namely the effective 3

4 isolated portions of envelope to be included in the simulation model, is defined to make the heat transfer analysis practical. In the slab (see Fig. 4), heat loss from hot water pipes (Q7) does not increase any more, if the calculation domain is extended beyond a certain limit. In the side wall (see Fig. 4), heat flow path becomes one dimensional beyond a certain limit at which there are no effects from the excessive heat loss at the thermal bridge. The effective calculation domains are extended to these limits in the slab and the side wall. The effective calculation domain will be varied according to the extent of insulation, outside and hot water conditions. For safety, the effective calculation domain is selected when it is most extended considering no subsidiary insulation, the lowest outside air, and hot water is supplied. Finally the effective calculation domain is determined by the steady state heat transfer analysis. Annual heating cost (C hjoint ) due to the annual heat loss through the joints (Q joint = Q4 + Q7) can be calculated by Eq. (3). C hjoint = Q joint UnitC ε fuel (3) The unit cost per MJ of the fuel (diesel) for heating (UnitC fuel ) is estimated to 6.84E-03$/MJ on the basis of the cost in Korea, The efficiency of the heating system in Korean apartment buildings (ε) is estimated to [6] THE OPTIMAL INSULATION DETAIL The side wall-slab joint is selected for developing the optimal insulation detail. Heat transfer analysis is performed with the finite difference multidimensional steady and unsteady state heat transfer analysis program, HEATrans v.8. [9] Thermal properties of all materials are assumed to be constant. Contact resistances at the interfaces between materials are neglected. In evaluating the inside surface condensation, a heat transfer coefficient of 34W/m 2 & is used on the outside surface. In calculating the annual heat loss, room is set to 23 &, hot water to 65&, sol-air s [1], and outside surface heat transfer coefficients applying the algorithm in DOE-2 [10] are used. Calculation time step is 15 minutes and the standard weather data of Seoul is used. Heat losses in July and August (nonheating periods) are excluded in calculating the annual heating cost. Each insulation detail alternative is assumed to have the same labor cost, since there is little difference in the amount of work and the process of construction. (see Fig. 6 and Table 3) Material and heating costs are calculated, when each alternative is constructed in the length of 9.2m, that is the length of side wall. Heat flow stream lines and the effective calculation domain in the side wall-slab joint are presented in Fig. 4. The lowest outside air of -15& in Seoul is used. Life cycle costs Life cycle costs of each insulation detail alternative (LCC joint ) are the sum of the initial costs (C ijoint = material cost + labor cost) and the present value of heating cost during the useful life time. Life cycle costs of each insulation detail alternative can be calculated by Eq. (4). LCC joint n ú ü (4) Chjoint 1+ r 1+ r 1 = Cijoint + r Real interests rate (r) and life time (n) are estimated to 2.27% and 30years, respectively. [7][8] Fig. 4. Effective calculation domain 4

5 ALTERNATIVES AND EVALUATION RESULTS Existing sectional insulation detail of the side wallslab joint is presented in Fig. 5. In this joint, region A and B are thermally defective areas. The insulation detail of region A adjacent to hot water pipes has a great influence on the heat loss. Firstly the optimal insulation detail of region A is determined without subsidiary insulation for region B. Then the optimal insulation detail of region B is determined while applying the optimal insulation detail of region A. Optimal insulation detail of region A Insulation detail alternatives for region A are presented in Fig According to the evaluation results (see Table 4), alternative 3 is the optimal insulation detail of region A. The lower part of the gypsum board is away from the concrete slab because of the difference between the gypsum board dimensions and the floor height. Optimal insulation detail of region B Insulation detail alternatives for region B are evaluated through 3 steps. (see Table 3) In step 1, alternatives are prepared to determine the optimal thickness of subsidiary insulation. (alternative 8 ~ 13) In step 2, alternatives are prepared to find the range where the optimal width of subsidiary insulation exists. (alternative 14 ~ 16) In step 3, alternatives are prepared to determine the optimal width of subsidiary insulation. (alternative 17 ~ 19) In step 1, it is found that the alternatives with the subsidiary insulation thickness of 10mm are more economical. (see Fig. 7) In step 2, it is found that the optimal width of subsidiary insulation exists between 200mm and 500mm. (see Fig. 8) In step 3, the life cycle costs are minimized at the subsidiary insulation width of mm from the regression analysis result. (see Fig. 9) The regression equation is y = E-07x E-04x E-01x E+03. (R 2 =9.9899E-01, F-value=1.9955E+03, P-value=1E-04) In the optimal insulation detail (see Fig. 10), the subsidiary insulation width of 300mm is adopted instead of mm for the convenience of construction and the economical use of insulation materials considering the manufactured size. 3 An external insulation method is not used due to the extremely high initial costs in Korea. The main insulation thickness is prescribed by law. Fig. 6. Alternatives for region A Table 3. Alternatives for region B No. of alternatives Subsidiary insulation dimensions (mm) Width No. of alternatives Subsidiary insulation dimensions Fig. 5. Existing sectional insulation detail Thickness Thickness (mm) Width

6 THE OPTIMAL INSULATION DETAIL The optimal insulation detail of the side wall-slab joint is presented in Fig. 10. Fig. 11 and Fig. 12 show the distribution of the and the standard inside air dew point (9.3&) in the joint during heating. The inside surface condensation does not occur under the standard condition. (see Fig. 12) The dimensionless lowest inside surface (T dl ) is Thus, it can be said that inside surface condensation never occurs in the optimal insulation detail in Seoul below an inside air relative humidity of 70% when the inside air is 20&. (see Fig. 2) Fig. 13 and Fig. 14 show the distribution of the and the standard inside air dew point (9.3&) in the joint while not heated. The inside surface condensation does not occur under the standard condition. (see Fig. 14) The dimensionless lowest inside surface (T dl ) is Thus, it can be said that inside surface condensation never occurs in the optimal insulation detail in Seoul below an inside air relative humidity of 63% when the inside air is 20&. (see Fig. 2) Evaluation results of each insulation detail alternative are presented in Table 4. The life cycle costs of alternative 3 are smaller by 11.76% than those of alternative 1 which has no subsidiary insulation. The life cycle costs of alternatives for region B are smaller by 12% than those of alternative 3 which has no subsidiary insulation in region B. Thus, the insulation detail of region A adjacent to hot water pipes is more important to reduce the life cycle costs. Life cycle costs($) Life cycle costs($) Life cycle costs($) 1,590 1,585 1,580 1,575 1,570 1,565 1,560 1,555 1,550 1,590 1,585 1,580 1,575 1,570 1,565 1,560 1,555 1,550 1,590 1,585 1,580 1,575 1,570 1,565 1,560 1,555 1,550 3 alternatives(thickness 5mm) alternatives(thickness 10mm) Width of subsidiary insulation(mm) Fig. 7. Life cycle costs trend in step Width of subsidiary insulation(mm) 12 alternatives(thickness 10mm) Range where the optimal width exists Fig. 8. Life cycle costs trend in step alternatives(thickness 10mm) Life cycle costs trend(thickness 10mm) Width of subsidiary insulation(mm) Fig. 9. Regression analysis result in step Even though the widths of subsidiary insulation for region B in alternatives 10 and 13 are wider than those in alternatives 9 and 12, heating cost increases on the contrary. This is resulted from the fact that the subsidiary insulation blocks the heat transfer from hot water pipes to the room in the lower floor. (see Q6 in Fig. 3). This means that heat loss increases on the contrary beyond a certain width of subsidiary insulation. Thus, the optimal subsidiary insulation width of 300mm should be applied to this joint. Fig. 10. The optimal insulation detail 6

7 Fig. 11. Temperature distribution during heating Fig. 14. Standard inside air dew point distribution in the joint while not heated Table 4. Evaluation results No. of alternatives Inside surface condensation while not heated (Ο : occur, : not occur) upper corner lower corner Initial (material) cost ($) Heating cost during life time ($) Life cycle costs ($) Life cycle costs variation rate (%) 1 Ο Ο , , , , , , A , , , , Fig. 12. Standard inside air dew point distribution in the joint during heating , , , , , , , , , , , , , , B , , , , , , , , , , , , , , Fig. 13. Temperature distribution while not heated 7

8 ACKNOWLEDGEMENTS This research was done as a part of the National Program for Developments of Energy Conservation Techniques. In this research project, 16 optimal insulation details for apartment building envelope were developed. The validity of the optimal insulation details was verified by the laboratory experiments. This research was financially supported by the Ministry of Trade, Industry and Energy (MOTIE) in Korea and Hyundai Industrial Development and Construction Co., Ltd. CONCLUSIONS The joints adjacent to hot water pipes in apartment building slabs become thermal bridges. To improve the thermal performance of these joints, the optimal insulation details should be applied after the evaluation of each insulation detail alternative. This study presents the methods for evaluating the thermal performance of insulation detail alternatives quantitatively using the multi-dimensional heat transfer simulation. The optimal insulation detail of the side wall-slab joint is presented based on the evaluation of inside surface condensation and life cycle costs. As results, thermal bridging effects are reduced and the life cycle costs are minimized. REFERENCES 1. ASHRAE, ASHRAE Handbook 1993 Fundamentals, ASHRAE, Silvers, J. P., R. P. Tye, D. L. Brownell, and S. E. Smith, A Survey of Building Envelope Thermal Anomalies and Assessment of Thermal Break Materials for Anomaly Correction, Vol. I - Survey and Assessment, Dynatech R/C Company for Oak Ridge National Laboratory, ORNL/Sub/ /1, Jul., Melton, B. S., P. Mulroney, T. Scott, and K. W. Childs, Building Envelope Thermal Anomaly Analysis, VVKR Inc. for Oak Ridge National Laboratory, ORNL/Sub/ /1, Dec., A, B, C Apartment Buildings Control Offices, Monthly Management Expenses Reports, Jan., 1996 ~ Dec., Childs, K. W., "Analysis of Seven Thermal Bridges Identified in a Commercial Building", ASHRAE Transactions, Vol.94, Korea Energy Management Corporation, Reports on the Results of Close Examinations of Apartment Buildings, Korea Energy Management Corporation, 1982 ~ Park, T. G., A Study on the Optimum Design Methodology Based on Life Cycle Cost for Domestic Apartment Houses, Ph.D. Thesis, Seoul National University, Dell'Isola, J. Alphonse, and S. J. Kirk, Life Cycle Costing for Design Professionals, McGraw-Hill Book Company, NewYork, Song, S. Y., and K. W. Kim, "A Study on the Grid System for the Computerized Multi-dimensional Finite Differential Heat Transfer Analysis Method", The Journal of Architectural Institute of Korea, Vol.11, No.4., Apr., LBL, DOE-2 Engineers Manual Version 2.1A, LBL, NOMENCLATURE T dl Dimensionless lowest inside surface T l Lowest inside surface (&) T o Outside air (&) T i Inside air (&) T c Inside air dew point (&) T oc Outside air at which the inside surface condensation occurs (&) Q1 heat gain on the outside surface Q2 heat loss from the stored heat in envelope Q3 heat gain on the inside surface Q4 heat loss on the inside surface Q5 heat loss on the outside surface Q6 heat gain from hot water pipes Q7 heat loss from hot water pipes C hjoint Annual heating cost ($) Q joint Annual heat loss (MJ) UnitC fuel Unit cost per MJ of the fuel for heating ($/MJ) ε Efficiency of the heating system LCC joint Life cycle costs ($) C ijoint Initial costs ($) r Real interests rate (%) n Life time of apartment buildings (year) 8